In my own experience, this has often been a good way of going in circles, where I end up undoing and redoing changes as fixing one thing breaks another, until I take a step back to find the proper algorithm by considering multiple inputs.

Of course, ymmv depending on how good your initial intuition is, and how tricky the problem is.

Again pytest makes things so much nicer in this regard. Having to comment things out sucks.

With pytest you can use the -x flag to stop after the first test failure.

Even better you can use that in combination with -lf to only run the last failed test.

> With pytest you can use the -x flag to stop after the first test failure.

> Even better you can use that in combination with -lf to only run the last failed test.

Fwiw `--sw` is much better for that specific use-case.

`--lf` is more useful to run the entire test suite, then re-run just the failed tests (of the entire suite). IIRC it can have some odd interactions with `-x` or `--maxfail`, because the strange things happen to the cached "selected set".

Though it may also be because I use xdist a fair bit, and the interaction of xdist with early interruptions (x, maxfail, ...) seems less than perfect.

Oh nice, didn't know about that flag!

Another option is to use a custom mark on the test you want to run, and then do something like "pytest -v -m onlyrunthis"

I think we also have to take into account how you work.

Are you working by yourself? Nobody else will ever see this code? Do whatever your timings tell you is best.

Are you working in a team or is this an open source project? In that case I think we need to do the thing that allows for the most variation. People that like having an overview of all the failures at once can only do so if you don't fail on the first error. If you do they have no workaround. If you show all of them though then your workaround is to simply only look at the first failure, ignoring the others.

Personally I do a hybrid approach. I want my test suite to continue through all tests and tell me everything that is wrong. If I change a small thing over here I definitely want to see whether I broke 3 or 4 unit tests around there or if I broke 37 test all around.

For fixing them I will go one by one after having done a quick survey of the types of failures. I. e. I might expect some of those failures due to the nature of my change. IDEs are really great here at least in my world. They allow me to very quickly run either a single test, a certain subset or the entire suite. After the one test I concentrated is green running all tests in that particular file again is just a click away. Now I am in overview mode again. I want to know what overall effect this one change had. Everything green again? Different subset broken? No change?